Sensoriamento remoto hiperespectral no estudo do corpo hídrico do baixo curso do rio Subaé, Bahia, Brasil (Hyperspectral remote sensing in the study of the body water of the low course of the Subaériver, Bahia, Brazil)

2018 
O sensoriamento remoto da Terra tem fornecido a base para mapeamento de atividades humanas e naturais, mostrando-se eficaz no monitoramento de sistemas hidricos, pois, os dados recobrem extensas areas e em diferentes faixas do espectro eletromagnetico, fornecendo informacoes sobre composicao de corpos hidricos. Desse modo o objetivo desse trabalho e analisar as variaveis espectrais de dados radiometricos obtidos in situ no corpo hidrico do rio Subae, Santo Amaro – Bahia – Brasil, com aplicacao de tecnicas de extracao informacoes de dados hiperespectrais relacionando a concentracao de clorofila-a como variavel limnologica. O periodo de amostragem seguiu meses chuvosos e secos. Os dados radiometricos foram coletados com uso de um espectrorradiometro FieldSpec 4 Hi-Res e a agua foi coletada para quantificacao da clorofila-a: os dados foram interpretados aplicando tecnicas de extracao de informacao sobre o espectro original na regiao entre 400 e 900 nm. Foram aplicadas a derivada espectral e a remocao do continuo observando feicoes caracteristicas da clorofila-a e a analise multivariada para correlacao da variavel limnologica com a variavel espectral. As tecnicas aplicadas foram eficientes na observacao das feicoes caracteristicas da clorofila-a, entretanto observou-se baixa correlacao da clorofila-a com os dados espectrais, alem de forte irregularidade nas curvas de 2a derivada. A espectrorradiometria tem grande potencial em estudos de corpos hidricos pois confronta metodos quimicos e fornece base para interpretacao de imagens orbitais, contudo ficou evidenciado que o comportamento espectral da agua e afetado por outros componentes opticamente ativos. A B S T R A C T Earth remote sensing have provided the basis for mapping human and natural activities, being effective in monitoring water systems, for data overlying extensive areas and in different bands of the electromagnetic spectrum, providing information on the composition of water bodies. Thus the aim of this study is to analyze the spectral variables radiometric data obtained in situ in the water of the river body Subae, Santo Amaro - Bahia - Brazil, with the hyperspectral data information extraction techniques relating to the concentration of chlorophyll-a as a variable limnological. The sampling period followed rainy and dry months. The radiometric data were collected using a FieldSpec 4 Hi-Res spectroradiometer and water was collected for chlorophyll-a quantification: the data that were interpreted by applying information extraction techniques on the original spectrum between 400 and 900 nm. The processing was carried out applying the spectral derivative and the continuum removal, observing features from chlorophyll, and the linear regression analysis was used to correlate the limnological variable with the spectral variable. The applied techniques were efficient in observing the characteristic features of chlorophyll-a, however, it was observed a low correlation of the chlorophyll-a with the spectral data, besides a strong irregularity in the 2nd derivative. Spectroradiometry has great potential in studies of water bodies as it confronts chemical methods and provides a basis for interpretation of orbital images. However, it has been evidenced that the spectral behavior of water is affected by other optically active components. Keywords: field spectroscopy, chlorophyll-a, spectral derivative, Subae river
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    0
    Citations
    NaN
    KQI
    []